UE4.5 released...

Epic designed the license on purpose to allow that. Unreal's income is primarily from the royalties, not the very minor subscription fees. If they didn't intend for it to be done that way, they would have never implemented that in the license.

 

Well then. I guess on the next full moon when I become slightly less lazy for about 5 minutes, I'll get myself a UE license to try out. If nothing else, it'll be a nice resume & conversation piece to not say "unity" for everything.

Any places Unity clearly outshines Unreal?

 

pro tip: 24 * $20 is still less than $600.

 

Unity is no match for Unreal in the 3D PC and Console market, it is years behind in toolsets and tech and most complex tasks can be done quickly with little effort. People keep saying "out the box", I dare most to try and match Unity with Unreal like for like. I wouldn't even attempt such a thing without a massive amount of staff and without the money to buy source, even though Unreal has 1/4 of the staff for every couple of months development = a year for Unity and they are falling forever further behind. https://trello.com/b/gHooNW9I/ue4-roadmap

Unity shines in mobile, 2D, platforms and stability, it's not a surprise though as Epic have been at it much longer. Have tons of experience in AAA development and heavy financial backing.

 

A poll for my interests, I have a few questions for those UE4 die-hards (ShadowK, Ryiah, etc), to answer with their honest opinions:

- Do you currently develop games in Unity or UE4? Or Both?
- Are you an independent developer? Or part of a bigger company?
- What types of games do you develop? Or are interested in developing in the future?
- What platforms do you typically target your games at?
- For Unity; What would be the 3 key features the engine needs to have implemented/redone?

 

Can I export/ compile my games for mac, linux and pc? That's my main concern. I don't really want to target the mobile market and I don't care if Unity is easier to use if unreal outperforms it on just about every task. I am half way through a class this semester that is nothing but SDL 1.2 and C!

 

Right. I was discussing a $60/month figure with a friend earlier which made it a slightly more expensive option than unity. I swear I can math!

 

I don't consider myself a diehard, I consider myself informed of my options. I pick the engine that is best suited for the task at hand. Even building one from scratch if that is a good option.

Currently on the designing stages for a mobile game with Unity.

Independent hobbyist who is looking to go professional. I may have the opportunity to form a company with some friends otherwise I am likely to stay independent.

As a hobbyist it helps to develop games I am interested in to stay motivated. I'm a big fan of roguelikes and even some arcade games.

PC. Planning to support Android soon (by buying a tablet during holiday sales).

Only one thing really stands out for me. A material editor. Shaders are a major aspect of graphics and are going to be used in practically every game. A visual node-based editor with the ability to preview at any node would be very nice. That's really the only feature from Unreal I truly miss while working in Unity.

 

1. We create games with both Unity and UE4, although that's being reviewed.
2. We are a company although new, we have a few subdivisions and one main branch. How we came to be is, I'm an ex engine developer that decided I wanted to make my own games.
3. Single player openworld RPG games, hack and slashes that sort of thing. One to be released in 2015 (hopefully), one in 2016..
4. PS4 / PC
5. Well, I'd need much more than three key features.. To keep it short what CryEngine / Frostbite / Unreal do.

Don't mistake me for a UE4 diehard, I'll use whatever tool equates to the highest amount of success as originally we tested CryEngine and various others. We even looked into building our own before the news of Unreal. I love the community here, as for tools I can take it or leave it. I've given feedback in various threads, I can collate them if you wish?

 

Interesting... I am purely asking to simply understand the backgrounds and environments that you work in. I apologise if I viewed you as potential UE4 die-hards (Although, you must admit, you can't stop talking about that engine in almost every single one of your posts. ).

 

And rightly so, in 2014 it's not only Unity that has competition. It's developers using the competition and if they are getting a leg up, then where does that leave people who invested time and money into Unity? We gave feedback and everything is still shrouded in mystery.. Unity should know the strengths of it's competitors.

In short, what are you doing about it? Can you do anything about it?

 

It gets brought up a lot in general on these forums. We just happen to be the most common responders.

 

My opinion does not count, just a hobbyist playing with toys. Still my 2 cents:

I don't want to be the heretic, but maybe Unity should not be (and should never have) trying to compete for AA and AAA customers. It's a brutal arms race that is not very compatible with wanting to run on older mobiles and slow desktops.

Be the best at what you do. That is small to medium Indies and hobbyists that target mobile or modest desktops. I think you should not fight to get ShadowK back, his current projects are too big.

Win those with ease of use and a nice package targeted at them. If they go later because they need more, so be it.

That target still needs stuff like:
- PBS (I create a room without textures and it looks awesome even it is programmer art, Indies dream)
- Node based shader editor
- terrain
- awesome examples are also very important (for guys like me that just play around and never finish anything)

Invest in 2D tools and mobile is my tip. In case Epic's Chinese Overlord wants best and fastest mobile 3D and 2D and pays for it, expect something later on. So be ready for that by preparing now.

 

@thxfoo

You might be right here, even in 2009 Unreal 3 was rocking a lot of the tools Unity is still to have. Like Kismet and material editor / 64-bit editor and cinematic tools. The trade off for cost of development might simply not be worth it, a shame because I really like Unity. I'd be happy if it was on par with UE3 (minus the horrible UI and coding portion)..

BUT! the only thing that gets me about that is one dude rocks up and in three months or so, by himself with limited knowledge goes and fills in for gaps the engine should already have. Pushes it out on the asset store... Sure why can't we do that ourselves? Of course we can do some of it. Not an issue, but that approach seems to follow with core components you can't touch. And why if they can do it? Why isn't Unity "doing it"?

I'd of probably never brought it up if not for Andy , not whining honestly .. Just trying to understand the position.

 

Those older devices do not last forever. What hardware were you running when Unreal Engine 3 emerged in 2006? I was using an Athlon XP @ 2.1GHz with 512MB memory and a GeForce FX 5700.

I was ecstatic when I first got it as it was my first brand new computer. I had mostly been using used parts before that with a Pentium 3 533MHz being the highest end. In retrospect though it was pretty bad.

It couldn't have run UE3 when it emerged because UE3 required shader model 3.0 and my FX 5700 only supported shader model 2.0a. These days though UE3 can run just fine off Intel HD, albeit at lower settings.

UE4's requirements may seem pretty steep now, but in 3 to 4 years those requirements will be pretty average.

 

Not sure, maybe:
- core code is a mess in Unity 4 so new features take long, adding external addon that must not work for all cases is easier
- Epic creates games (e.g. they have 5 different reflection methods, they saw they are needed from using it and seeing where which fails)
- Epic has some exceptional hero coders (it's like trying to catch up to ID in Carmack's time, till you catch up he had a new engine destroying everything that came before). I don't say Unity does not have such heroes, but Epic seems to have more.
- Of course Unity should have bought ShaderForge and such stuff, but before UE4 there was no need (to keep Indy customers, AAA where elsewhere anyway). Extra income from Store that way.
- Unity can do much that is in the Store now. But most of that stuff does not have the UE4 level. But more than good enough for most Indies. Still seamless integration is a lot of extra work.

@Ryiah: I don't mean Unity should do nothing now. Even staying relevant for Indies will be a lot of work. Pick stuff that Indies want in a smart way. Begin implementing current AAA stuff, but till you have it the AAA train will be at the next thing. But don't waste all resources on that, it can take some years delay for most features, better use the rest of the resources to make what you have perfect and complete. Indies are happy just having the previous AAA thing as long as it is stable, easy to use and complete.

 

UE3 2006 is a good point. As ShadowK said, he would be happy to have that now in Unity. So in this case they could have had 8 years to catch up and the Indy customer would still be happy. And being late has a huge advantage, the implementation could have much less performance, because they get away with doing on today's computers what UE3 did on those of 8 years ago.

Edit: and you can look at the iterations the AAA engine did to get there. So you directly know what works. Saves money and time. With Steve Jobs/Picasso: 'good artists copy; great artists steal'

 

Spot on, it's not like the engine aged badly either.. The new Arkham Knight is UE3 with some lighting tweaks..

 

Yeah I know, it took them a total of 5 months to get to this point... sheesh /sarcasm You do know that you could generate lightmap UVs inside UE4 since release, right? They are just making it an automatic option. Do you know how long it took Unity to get lightmapping???? I've been using Unity since 2.x, and it wasn't until 3.0 that we got lightmapping (a little more than a year).

 

I'm pretty sure the visual material editor has been on the top of all these comparison lists since they began(for good reason). Although I am having way too much fun with their event driven behaviour tree editor as well.

 

Yes, I had been using the manual option on each 3D asset in UE4 versions prior to UE4.5. I am well aware of the feature. That is why the automatic option is so appealing. Honestly though, the hot reload for C++ code was a much bigger deal for me than the lightmap UV.

 

I have been rather critical of Unity3D in the past. Mostly because I am really frustrated with the Unity3D Tool set. So here is my 2 cents.

1. I develop games using both. I am using Unity 3D in my spare time right now. Unreal 4 gets most of my attention however due to its blueprints system and Material Editor. I want to finish my little futuristic GTA game with Unity3D.
2. Independent Developer
3. Currently I am developing a Board game with Unity for personal growth with C# and an unannounced game with UE4.
4. PC (Windows, MacOSX, Linux) I would like to target SteamOS if possible.
5. Personally - I think the following should be done.

• Make Unity more Designer/Artist friendly that should be a STRONG focus for Unity Technologies. Things like a visual logic editor (Similar to Blueprints or Kismet) and a official visual Shader Editor. Would be a step in the right direction.
• Re-evaluate the feature set for Unity Free - I am really frustrated with how Unity Tech intentionally gimps Unity free every major release. I am sure i am not the only one. I look at UE4 and see everything that I can get for 19.00 USD a month. I look at Unity Free and the offering is lacking. My advice - Do not limit your free customers by feature set. Give your users everything and they will make awesome stuff in return which will make more people want to use Unity. Even if that has a small upfront cost like five dollars or something.
• Better support for Vehicles. Quadarpeds, hexapeds, etc. in Mecanim.

That is my 2 cents - You asked so I gave you my honest answer.

- HeadClot

 

Since I'm using Unity since... gosh... 2.5 or so...

1. No games but VR applications using both (Unity for mobile and partly Oculus testing, UE Oculus); games in the planning
2. Independent with 3 employees, working upon customer request.
3. architectural VR, product presentations, interactive experiences; game planning -> strategy
4. PC and mobile (Android mainly; easier to distribute to)
5. Different stuff (that don't seem to get "fixed" with 5.x):

- Integrate more of the tools that are actually needed into Unity instead of relying on Asset Store producers (i.e. Shader creation, Visual Scripting, ...)
- Improved terrain engine
- Make all pro licenses one single pro license (it really buggers me that I have to pay twice as much only to be able to deploy to Android)

There are other things that need to be "revisited", but that is mostly candy for my case...

There should be a 6th question: "How to Unity Technology have to improve as a company?"

Answer: Actually, you seem to go into the right direction. While the weekly patch updates start to look good, it overall product cycle is much too long. You currently have 160 developers (or overall employee count?=. The feature/bugfix output for so many developers is much too low. The iteration between .x versions is too long, especially if new features are announced "somewhere in the .x cycle" (which basically means somewhere between 6 months and 2 years).

I understand that it's no easy task to implement, maintain and improve an engine such as Unity (been there, done that), but seen from the outside the speed at which things improve with Unity is sub-par compared to competitors.

Unity still has big advantages for everything mobile related (main reason we stick with Unity for our customer work), but as a german wording says "Du sollst dich nicht auf deinen Lorbeeren ausruhen." (means you should not rest on your laurels). 5.0 looks interesting, but with any major Unity update, it will only be save to do professional work somewhere in .x versions (one the .0 bugs are sorted out and fixed) (and no, I'm not going to enumerate the problems encountered in 3.0, 4.0 releases) (As a sidenote: for some mobile work we still stick with 4.3.4 because of regressions we encountered in later versions and which haven't yet been fixed (at least I couldn't find the bug id related to a fix)).

 

This has been brought up several times in the past, but I too want to bump this issue: the issue of asset store. I've seen a few times that when someone asks for certain functionality, UT devs come into the thread and say "hey, there's a solution on the asset store, why don't you use it?".

I think this is one of the issues with Unity. If there is a good solution on the asset store, the internal implementation might not be prioritized enough. I don't know exactly if that is the case, but it does seem logical to me. Not only can they work on something else, they are also getting paid for not having this functionality built-in.

And as a side note, not everyone can afford some of the assets on the store.
I wish UT would try buying out some of the assets, such as Shader Forge, Text Mesh Pro, etc., and merge them into Unity.

 

I understand that it's not trivial. But let's take the example of Text Mesh Pro and distance field text rendering. A working solution is right there, on the asset store, with lots of functionality. Will this ever be built-in into Unity and is it even on the roadmap right now? How long will it take before it's implemented?

 

Tim C developed the stumpy shader editor back in 2010. That's 4 years you had time to integrate it into Unity.

Sorry Aras, but it certainly looks like someone said:"Tim, drop it, there is shader forge and 30% for us from the price."

 

Please not THAT way. Bugs and fixing is and stays a dark chapter in Unity. The current open Beta is a good step in the right direction. But can be improved.

I wait for example since weeks for the input field to become stable enough to work with. Every new release breaks it even more it seems. And with the new version just arrived i already wait for the next version because my show is stopped again. I didn't got a workaround told.

When you want to see a great bug fixing then have a look how Blender does it for example. They do it really well. The most interesting feature is the Daily builds with the newest fixes. And you don't need to VOTE for your bug, EVERY bug gets fixed when reproduceable and catchable. The bug list is open and everybody can scroll through and have a look if its bug is already on the list or not.

From what i heard Unreal handles it equal fast and nice.

 

^THIS is still my main issue with Unity. Personally I do mobile games ONLY, and yet to see the reason why I have to buy pro for PC PLUS iOS and Android?!

I mean as soon as Epic catches up in the mobile sector it's gonna be tough for you guys.

 

@Andy Touch,
- Do you currently develop games in Unity or UE4? Or Both?
First, to let you know my standpoint, I'm not on anyones side, I'm by no means an UE4 (neither Unity) fan. We use both of those engines. ^^
- Are you an independent developer? Or part of a bigger company?
I was part of bigs companies (Ubisoft/Gameloft and other softwares companies in Montreal) and funded my own one in 2009, so currently Independent Studio working on two projects.
- What types of games do you develop? Or are interested in developing in the future?
Adventure/horror games, with large open streamable worlds.
- What platforms do you typically target your games at?
Mobile, PC, Consoles.
- For Unity; What would be the 3 key features the engine needs to have implemented/redone?
-A complete real-time workflow, rendering/culling. Even Mobile hardware is approaching the consoles/pc performance, plus their APIs are making the process even faster (iOS Metal).
-Streamable friendly (better/faster/easier support for large-worlds).
-Terrain system with river/road tools and why not a fully featured ocean system.
- Bonus: a TimeOfDay system and an Ubershader (a real-one that can output any kind of surface material such as skin, hair, etc).
In other words, Unity with some awesome CryEngine features. ^^

 

Im a new user in the forums and what i know about unity is from paper and youtube videos(until ver 5 comes out which i plan on using)so i hope that i dont say something stupid.If i do,please ignore this post.
I realy think that the future unity shoud have adleast a way to preview/play the unfinished ios game in widows environment and once finished than move it to the mac for the rest.As technology advances people shoud spend more time with their assets,programing and game mechanics rather than moving from one place to another in order to see the results.Its a waste of time.In the end if a user wants to create a game for ios he will eventualy buy a mac.Apples work is done here so i dont think that thigs need to be more dificult for the developer than what they are.Im sure the unity team can make this the same way epic did in udk years back.
On the other hand from what i see once you start unity you start with a emty scene.Its fine for me but adding a window that has some basic templates and very basic game mechanics like a fps option ,a side scroller,isometric or basic things like that for new people to have a choice and start with something woud be i good idea and motivating as they will see some instant result.
And the price...I remember when engines were so expensive and some of them are still but time changes and companyes see that there are many developers.I think that the current unity price is high,apart from that you have to buy ios and android separatly.For 1500 adleast the ios and android packs can come free for the pro version.

Its a personal opinion and i hope that is ok to say it.

 

Why was it so difficult? I mean the new UI ended up being exactly like NGUI, so why did it take years to implement/integrate what's been available for years on the asset store?

Text Mesh Pro, shader forge, marmoset, baking terrain to mesh, and a goddamn ramp brush for the terrain editor!

 

You know you can already make your own custom brushes? Talked about here, but seems to be undocumented.

However I'd really like to se pressure sensitivity for terrain painting. There's already a pressure event (Event.current.pressure), again undocumented, so it should simply be a matter of multiplying the terrain modifications with this.

 

Well Text Mesh Pro is kind of rock bottom on priority. It's hardly difficult to fork out $60 and support the developer. It's not an asset that would get any quicker really by being integrated.

But what would really smash things into orbit is what people have been asking for, for a very long time and what UE3 has had since 2006 - a node based material editor. The reason this is so important to be done by Unity is because of how well optimised it could be. Currently, shader forge is nice, but hardly optimised to the level where its a great idea for mobile. Unity could change that.

It's things that are pretty obvious like a built in node based editor and blueprints that attract artists in droves to UE4 - they feel empowered and capable to do it without a programmer so I would consider these two things to rank really high on priority for Unity after the required U5 engine stuff.

 

Well the author has little interest in mobile optimisation paths (he said so) and Unity has a wealth of experience in shader optimisation. So I think Unity would change that.

 

Well, that's because Unity is a really comfortable engine to work with (things like customizing the UI is a bliss), except for the part it hasn't still even matched Unreal Engine 3 in what the engine is able to output graphically.
Some keep saying "Unreal Engine is for AAA devs with big budget". Well, I have seen anything done in Unreal Engine and CryEngine done in Unity. Problem starts when we start factoring graphics, performance or anything rendering for that matter... (and certain animation annoyances).
Add in the lack of certain tools and features that come built-in with the other engines and you start wondering if that "commodity" is worth it. As many useful features as there are in the asset store, the sentiment "It isn't integrated into the engine therefore it isn't as good as it could be" will always be there.

If you have read before, Shadowk managed a good looking game in Unity. His problem is that he just ran out of places in where to optimize and in where to cut without compromising his product/game. Of course, we have been counting in Unity 5 fixing most of them, but the second you said "Coming later in the 5.x" cycle, we started associating it with "UI coming later in the 4.x" cycle... and here's when you start to notice how not keeping your words comes later to ~that word~ in the ass.

All Unity has to do is matching CryEngine in the rendering and performance department and you will see how most request/complains from those doing 3D games die down. (unlikely, but you should set yourselves that high of a goal so even if you fall short, you have something good... and yes, I am aware of how shameless "match CryEngine" sounds, but I am sure Unity itself didn't start aiming nowhere, the bar has just been upped)
Right now it stands as:
CryEngine -> "Not even CryteK understands it" easy to work with (For no reason. Cut the "hard to work with = more powerful" lie, please), amazing rendering and graphical output, godly performance.
UE4 -> Mid difficulty to work with (easy once proficient), decent rendering (standards vary per person, I just call it "UE3 with a few new shiny features"), average performance (for now).
Unity -> Easy to work with from the beginning, difficulty outputting nice looking graphics even with high quality assets and even then, when you aim high, bad performance without even reaching the other ones output. [See: The Forest]

What got me interested in game design was CryEngine, and now I am rooting for Unity (mostly have to do with me not liking "the unreal way" even when it was the first thing I used). That should tell you something.
...
Though that "rooting for Unity" has mostly to do with me being able to wait because right now even though I am building 3D levels and assets, and planning things for a game in the future, I am already waiting for Unity 6. All to do with my current focus is 2D games so "LOOK! A GI that doesn't crash Windows!" isn't still a current concern for me.

Also, I keep reading "Unreal Engine 4 source is open" as "Say goodbye to engine updates". I am sure I can't be the only one.

 

Well, that's exactly what I'm talking about. It's a small feature (the same kind of small features are added in bunches in every UE update), but it will never be added into Unity. Because it's on the asset store, "it's hardly difficult to fork out $60 for", and it won't benefit from improved performance if integrated.
P.S. Where I live, average salary is ~$300, so forking $60 for this asset is not exactly a knee-jerk reaction.

 

But it is hard to fork out 1000 x $60 for all the missing functionality that should be in the trunk. But is just available as a plugin.

I do of course understand that the addon developers wants to live too. That's what makes it so difficult for Unity. To put functionality from available plugins into the trunk afterwards means the addon developer is loosing its income. Happens at the moment with the UI solutions. On the other hand, it is a competition problem for Unity that you need to buy core functionality from plugin developers instead of getting it natively.

And when that core functionality would've been in Unity in the first place, then there wouldn't have been the plugins available. Means those developers would have investigated into other stuff.

 

Well you guys are saying you don't want to buy on asset store, but just want it for free which means pro owners are paying for it, and it shouldn't be in Unity Free.

Think about it...

 

I'm a pro owner

 

Nothing to do with wanting it for free in the first place. It's about having it natively in the trunk, and not external. Every external extension has the danger of not working with the next update or patch anymore. And dealing with several extensions and the trouble they brings makes your own projects harder to maintain. I just hated the hassle which the handful of plugins that i use made to me with every update.

Of course it is also a thing of money. I pay 19 dollar once to have UE4. And have everything needed, because everything needed is in the trunk. In Unity i pay 1500 for pro, plus 60 here plus 50 there, plus 200 there, plus 20 there, and there, and there, and there ... for stuff that should be in the trunk.

I think, or better know, that you have stuff in the asset store. And you are of coures not willing to loose that income. That's something i can understand. But you have to understand that core functionality should not be delivered by extension developers. Core functionality should be available by the main software.

 

I do not have any assets in asset store nor do I plan any.

 

Whoops, you are right, i did mix you with another moderator here, sorry.

 

@Frpmta

I was going to write a long winded post on the hackathon that ensued dragging Unity screaming to be a challenger so were not 5 years behind even out the gate, we spent all our time on GFX because with DA3 and Witcher 3 coming up we had no choice. But no point, I understand Unity has a lot going on even with the shear amount of platforms they have to support and sorting a replacement for Mono.

But as our customers care little for how we do it, the same applies for our tools (Although I'd like to believe were a lot more understanding).

Which brings me back to the age old question and one that would Enlighten (NPI) all of us, what is Unity's focal point? 2D / 3D / mobile? I wouldn't count my chickens in the mobile market with tech like Tegra coming out, Unreal can run flat out on it as long as the scene has static baked lighting.

 

I could resume in some points the pros and cons :
Mobile : Unity
UE4 is too limited for Android, only few paltforms are supported , even my simple 100$ Android phone able to run many 3D games was not supported. Epic is concentrating on high powered mobile for big graphics insteadof doing like Unity to optimize 3D and have good frame rate on more devices.

Easy to use tools and programming : Unity
It's the easiest framework and easy to take and learn, and C# you can use it without paying anything more.

Tools : UE4
About a shader editor , modeling tools , visual programming , Unity should integrate some like shader forge, Probuilder, Playmker or others ones . UDK 3 have these tools from lot of years now, Unity shoud not stay behind and have them out of the box. As we have them out of the box for some 20$ today.

Releases : UE4
Integration of new things take a lot of time in Unity, why? because of only one or two guys working on a features instead of having a team sharing the same project ?
It took so long time for Unity to bring a good UI, and it's Beta , while Epic realeases many new things and some major rendering/Animation features in some months.

Big levels : UE4
About UT5 it will have to proove itself for the ability to run bigger games, levels , streaming , multi threading perhaps , good lightening in big outdoors, only future games will tell.
I don' t say UE4 is best for big outdoors but it is more designed to handle bigger ones

Last graphics for lower price : UE4
If you have 20$ and want to make a desktop game and you want to add to your game PBR, advanced particles, advanced lightening and other full screen effects , UE4 is a better deal than Unity, because it's lot more cheaper and because it is simply available.

I work with UE4 as 3D artist and blueprint in a indie team project ; and i use personnaly UT4.6 that i enjoy a lot ,targetting 3D mobile and some 3D small desktop indie games. I love both engine, each for their own strong points.

Each tool is better some areas than another tool

 

Or being unavailable. Just look at the relatively recent incident with Daikon Forge and how it had to be hacked into being available to those who purchased it.

 

How many of those devices are going to still be in highly active use in three to four years? Epic is simply aiming their engine towards the future rather than trying to support the present. Unity will support the devices that exist now, but Epic will be able to take advantage of the hardware when consumers are finally using that level of hardware.

No, it takes them a long time because they're maintaining backwards compatibility. Epic throws that out the window with each new major release of their engine. It visibly shows because they both are lacking some features from the previous generation but are implementing functionality at an absurd pace.

This only holds true as long as the tools have a roughly similar speed of improvement. Unity is obviously very slow at improving and Unreal is very fast. Unreal could potentially catch up on the mobile front and kill off the advantage.

 

Are you aware of so much people not interested hard core games in mobiles, only buying for communication first, games in second, and playing some games from time to time.
There is so much great 2D games (RPG, beat them up, adventure etc ...) , 2D is possible on very large hardware even within 5 years. While Unreal targets last hardware mobiles with best possible 3D.

It's also a choice Epic makes , but customers exists in many realms , not only last hight tech phones . So like many people i won't choose UE4 for mobile even on for 3D mobile games as we don't target last graphics and techs phones.

This is not speed of improvement in some cases that matters, but the project goals and needs.
If you make 2D games you don't need tons of releases, if you target simple forward rendering 3D games with some indie style, in that case your core will be gameplay and not last PBR or lightening or physic cloth system.
If you target a game that shows a lot visually and with physics and particles, cinematics, yes in that case UE4 is the tool.
The right tool for your personnal project goal and needs.